When users of a conventional information retrieval system search for information about a feature using, for example, a keyword search, a system may return a list of pages that include the keywords and provide the list of pages in a sequence according to rank scores of the pages. The system may generate the rank scores based on any of a number of criteria including inferred opinion and structure of the documents. For example, a network's link structure can be used as an indicator of an individual page's value. A link from page A to page B can be interpreted as a vote, by page A, for page B. Votes cast by pages that are themselves important weigh more heavily and help to make other pages important. The page can also be ranked based on locations and frequencies of keywords within the page. For example, if the keywords are located in the HTML title tag of the page, or if the keywords appear more frequently in the page, the rank score of the page may become higher. The usefulness of the search can depend in part on the quality of the rank scores, such as whether more relevant pages are ranked higher than less relevant pages.